The author has an agenda, and the derogatory part about DL papers in general is unconvincing. However, the part where the author replicates the results of the Uber CoordConv without having to append coordinates to the output is nice. What do you think?

Ps this sub already identified the Uber AI paper as lacking substance, a few weeks ago.



Source link
thanks you RSS link
( https://www.reddit.com/r/MachineLearning/comments/90n40l/_of_a___paper_quite_brutal/)

LEAVE A REPLY

Please enter your comment!
Please enter your name here